Z-flying focal spot CT reconstruction without dataset combination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current z-flying focal spot (zFFS) reconstruction algorithms for computed tomography (CT) systems face limitations in field-of-view (FOV) and accuracy, particularly for voxels outside the limited FOV, due to assumptions about ray stacking and geometric spacing, leading to inaccuracies and artifacts in image reconstruction.
Innovation Solution
A new reconstruction method that acquires and processes datasets from two focal spots separately in fan geometry without combining them into a single geometry, using a weighted re-binning and back-projection algorithm to maintain native geometry and achieve equal spacing of rays along the z-axis, allowing for improved image reconstruction beyond the limited FOV.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If datasets from two focal spots are combined into one dataset with interleaved sampling, then the field of view is limited to approximately 200 mm, but the reconstruction can be performed using regular single focal spot geometry
Solution Approach 1:
The patent divides the reconstruction process into separate handling of datasets from different focal spots. Instead of combining all data into a single interleaved dataset, the method processes each focal spot's data independently through separate re-binning operations, then combines the reconstructed images. This segmentation avoids the FOV limitations and geometric assumptions required by interleaved sampling methods.
Solution Approach 2:
The patent transitions from combining data in the z-dimension (interleaved sampling along the axial direction) to combining data in the image space dimension after separate reconstructions. By performing independent reconstructions for each focal spot and then combining the resulting images, the method operates in a different dimensional space, avoiding the geometric constraints of z-dimension interleaving.
2Ease of operation
If interleaved sampling is used to combine datasets from alternating focal spots, then the data can be processed with regular single focal spot geometry, but geometric spacing assumptions are violated for voxels outside the limited FOV
Solution Approach 1:
The patent segments the reconstruction process into independent stages for each focal spot. Each dataset is re-binned and reconstructed separately using appropriate geometric corrections for its specific focal spot position, rather than forcing all data into a single geometric model. This maintains operational simplicity while improving accuracy through targeted geometric handling.
Solution Approach 2:
The patent applies local quality by using focal-spot-specific re-binning parameters and geometric corrections tailored to each focal spot's unique position and characteristics. Each dataset undergoes processing optimized for its particular geometry, rather than applying a uniform geometric model to all data, thereby maintaining accuracy across different FOV regions.
3Manufacturing precision
If zFFS strategy is used to increase sampling rate in z-direction, then axial resolution and z-direction sampling are improved, but image reconstruction artifacts increase with conventional interleaved methods
Solution Approach 1:
The patent segments the zFFS data processing into separate reconstruction streams for each focal spot. By independently re-binning and reconstructing data from each focal spot using geometry appropriate to that specific focal position, the method preserves the high axial resolution benefits of zFFS while avoiding the artifacts that arise from incorrect geometric assumptions in interleaved reconstruction methods.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The proposed method enhances image reconstruction accuracy and reduces artifacts, enabling imaging of voxels beyond the previous FOV limitations, as validated with offset zFFS scans of a physical head phantom, and maintains image quality for larger fields of view.
Implementation Method 1
the x-ray tube generates high speed electrons from the filament. The electrons fly toward the positive target anode, in which the energy of the electrons is converted to X-rays
Implementation Method 2
the scintillation crystal absorbs x-rays and converts the absorbed energy into visible light
Implementation Method 3
A photodiode is used to convert the light to an electric current
Data Source
AI summary
A computed tomography (CT) system includes a rotatable gantry having an opening to receive an object to be scanned, an x-ray tube having an anode, the x-ray tube positioned on the rotatable gantry to generate x-rays from a first focal spot at a first z-location, and from a second focal spot at a second z-location, a pixelated detector positioned on the rotatable gantry to receive the x-rays from the first z-location and from the second z-location, and a computer. The computer is programmed to acquire a first dataset in a fan geometry at a first z-location, acquire a second dataset in the fan geometry at a second z-location, and reconstruct an image based on the first dataset and the second dataset, wherein the reconstruction is performed without combining the first dataset and the second dataset into one dataset with a single geometry from which the image reconstruction is performed.